http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : Niyazi K l ç (검색결과 1 건)
- 무료
- 기관 내 무료
- 유료
-
An Improved Hybrid Feature Reduction for Increased Breast Cancer Diagnostic Performance
Ahmet Mert,Niyazi K l ç,Ayd n Akan 대한의용생체공학회 2014 Biomedical Engineering Letters (BMEL) Vol.4 No.3
AbstractPurpose Early and correct diagnosis of a disease is vital forthe success of treatment. Medical diagnostic decision supportsystem can be used to improve the accuracy of the traditionaldiagnosis. As such, various pattern recognition methods arestudied and applied to develop medical diagnostic decisionsupport system. In this study, the effects of dimensionalityreduction techniques with probabilistic neural network (PNN)on breast cancer classification are examined. Methods A hybrid method is proposed using the independentcomponent analysis (ICA) and the discrete wavelet transform(DWT) to reduce feature vectors of Wisconsin diagnosticbreast cancer (WDBC) data set. Two independent components(ICs), and one approximation coefficient of the DWT areused as a reduced feature vector to classify breast cancer usingPNN. Performance measures such as accuracy, sensitivity,specificity, Youden’s index and the receiver operatingcharacteristics (ROC) curve are computed to indicate theadvantages of the hybrid feature reduction. Results The proposed feature reduction approach using ICAand DWT improves the diagnostic capability of the PNNclassifier. The hybrid feature reduction has a higher diagnosticcapability than the original thirty features using PNN as aclassifier. Accuracy and sensitivity are 96.31% and 98.88%,while the results of the classification using the original thirtyfeatures are 92.09% and 95.52%. Conclusions Feature reduction approach using ICA andDWT together increases the performance measures of breast cancer classification using PNN, while reducing computational complexity.
연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
